All Versions
24
Latest Version
Avg Release Cycle
87 days
Latest Release
1586 days ago

Changelog History
Page 2

  • v0.10.0 Changes

    March 14, 2016
    • 🖐 Handle None inputs same as empty string (Issue #94) [Nick Miller]

  • v0.9.0 Changes

    March 07, 2016
    • ⚡️ Pull down all keys when updating local copy. [Jose Diaz-Gonzalez]

  • v0.8.2 Changes

    February 26, 2016
    • ✂ Remove the warning for "slow" sequence matcher on PyPy. [Julian Berman]

    where it's preferable to use the pure-python implementation.

  • v0.8.1 Changes

    January 25, 2016

    • 🚀 Minor release changes. [Jose Diaz-Gonzalez]

    • Clean up wiki link in readme. [Ewan Oglethorpe]

  • v0.8.0 Changes

    November 16, 2015

    • Refer to Levenshtein distance in readme. Closes #88. [Jose Diaz- Gonzalez]

    • ➕ Added install step for travis to have pep8 available. [Pedro Rodrigues]

    • ➕ Added a pep8 test. The way I add the error 501 to the ignore tuple is probably wrong but from the docs and source code of pep8 I could not find any other way. [Pedro Rodrigues]

    I also went ahead and removed the pep8 call from the release file.

    • ➕ Added python 3.5, pypy, and ypyp3 to the travis config file. [Pedro Rodrigues]

    • ➕ Added another step to the release file to run the tests before releasing. [Pedro Rodrigues]

    • 🛠 Fixed a few pep8 errors Added a verification step in the release automation file. This step should probably be somewhere at git level. [Pedro Rodrigues]

    • Pep8. [Pedro Rodrigues]

    • Leaving TODOs in the code was never a good idea. [Pedro Rodrigues]

    • 🔄 Changed return values to be rounded integers. [Pedro Rodrigues]

    • ➕ Added a test with the recovered data file. [Pedro Rodrigues]

    • Recovered titledata.csv. [Pedro Rodrigues]

    • 🚚 Move extract test methods into the process test. [Shale Craig]

    Somehow, they ended up in the RatioTest, despite asserting that the ProcessTest works.

  • v0.7.0 Changes

    October 02, 2015

    • ✅ Use portable syntax for catching exception on tests. [Luis Madrigal]

    • ✅ [Fix] test against correct variable. [Luis Madrigal]

    • ➕ Add unit tests for validator decorators. [Luis Madrigal]

    • 🚚 Move validators to decorator functions. [Luis Madrigal]

    This allows easier composition and IMO makes the functions more readable

    • 🛠 Fix typo: dictionery -> dictionary. [shale]

    • FizzyWuzzy -> FuzzyWuzzy typo correction. [shale]

    • ➕ Add check for gitchangelog. [Jose Diaz-Gonzalez]

  • v0.6.2 Changes

    September 03, 2015
    • 👕 Ensure the rst-lint binary is available. [Jose Diaz-Gonzalez]

  • v0.6.1 Changes

    August 07, 2015
    • Minor whitespace changes for PEP8. [Jose Diaz-Gonzalez]

  • v0.6.0 Changes

    July 20, 2015

    • ➕ Added link to a java port. [Andriy Burkov]

    • Patched "name 'unicode' is not defined" python3. [Carlos Garay]

    https://github.com/seatgeek/fuzzywuzzy/issues/80

    • 👉 Make process.extract accept {dict, list}-like choices. [Nathan Typanski]

    Previously, process.extract expected lists or dictionaries, and tested this with isinstance() calls. In keeping with the spirit of Python (duck typing and all that), this change enables one to use extract() on any dict-like object for dict-like results, or any list-like object for list-like results.

    So now we can (and, indeed, I've added tests for these uses) call extract() on things like:

    • a generator of strings ("any iterable")
    • a UserDict
    • custom user-made classes that "look like" dicts (or, really, anything with a .items() method that behaves like a dict)
    • plain old lists and dicts

    The behavior is exactly the same for previous use cases of lists-and-dicts.

    This change goes along nicely with PR #68, since those docs suggest dict-like behavior is valid, and this change makes that true.

    • 🔀 Merge conflict. [Adam Cohen]

    • 👌 Improve docs for fuzzywuzzy.process. [Nathan Typanski]

    The documentation for this module was dated and sometimes inaccurate. This overhauls the docs to accurately describe the current module, including detailing optional arguments that were not previously explained - e.g., limit argument to extract().

    This change follows the Google Python Style Guide, which may be found at:

    https://google-styleguide.googlecode.com/svn/trunk/pyguide.html?showone=Comments#Comments

  • v0.5.0 Changes

    February 04, 2015

    • 🛠 FIX: 0.4.0 is released, no need to specify 0.3.1 in README. [Josh Warner (Mac)]

    • 🛠 Fixed a small typo. [Rostislav Semenov]

    • 0️⃣ Reset processor and scorer defaults to None with argument checking. [foxxyz]

    • Catch generators without lengths. [Jeremiah Lowin]

    • 🛠 Fixed python3 issue and deprecated assertion method. [foxxyz]

    • 🛠 Fixed some docstrings, typos, python3 string method compatibility, some errors that crept in during rebase. [foxxyz]

    • [mod] The lamdba in extract is not needed. [Olivier Le Thanh Duong]

    [mod] Pass directly the defaults functions in the args

    [mod] itertools.takewhile() can handle empty list just fine no need to test for it

    [mod] Shorten extractOne by removing double if

    [mod] Use a list comprehention in extract()

    [mod] Autopep8 on process.py

    [doc] Document make_type_consistent

    [mod] bad_chars shortened

    [enh] Move regex compilation outside the method, otherwhise we don't get the benefit from it

    [mod] Don't need all the blah just to redefine method from string module

    [mod] Remove unused import

    [mod] Autopep8 on string_processing.py

    [mod] Rewrote asciidammit without recursion to make it more readable

    [mod] Autopep8 on utils.py

    [mod] Remove unused import

    [doc] Add some doc to fuzz.py

    [mod] Move the code to sort string in a separate function

    [doc] Docstrings for WRatio, UWRatio

    • ➕ Add note on which package to install. Closes #67. [Jose Diaz-Gonzalez]